Unexpected benefit of sorbitol placebo in Mg intervention study of premenstrual symptoms: implications for choice of placebo in RCTs

Abstract

We carried out a randomized, double-blind, crossover study of 85 women, designed to investigate the dose-response of daily Mg supplementation on premenstrual symptoms. Each woman took one of four treatments: Mg (200, 350 or 500 mg/day) or sorbitol (placebo) for 2 months. This was followed by a washout of 1 month, and then each woman received one of the three remaining treatments for a further 2 months. Unexpectedly, sorbitol (1305 mg) reduced anxiety-related and total premenstrual symptoms after 2 months compared with Mg treatments (P<0.001 and P<0.001, respectively). We conclude that low-dose sorbitol reduces premenstrual symptoms beyond that expected of a placebo. After 2 months of treatment, sorbitol also reduced urinary Mg excretion compared to baseline (no intervention) and Mg treatments (P=0.005). A follow-up study on 17 healthy volunteers confirmed lack of effect on urinary Mg output of a similar sorbitol intervention regime compared with either baseline or cellulose placebo. It appears that sorbitol may influence Mg homeostasis in women suffering premenstrual symptoms, but not in healthy individuals. Implications for placebo choice in RCTs are discussed.